Two-way communication networks typically employ a transactional communication protocol. Under such a protocol, one connects to something, makes a request, and one receives a response. TCP/IP, SQL, and http all use this principle. Even file access across networks is based upon this principle. Two-way networks face some fundamental problems, among them, how a content source responds to a sudden increase, known hereafter as a spike, in requests to that source. One drawback of the standard approaches to solving this problem is that they render the content source inoperable when the requests pass the content source's response limits.
The development and deployment of residential broadband systems has enabled millions of users to receive hundreds of television channels. Thus, there are known successful merges of at least partial two-way communications and residential broadband delivery, such as the Advanced Television Enhancement Forum (ATVEF).
All of these systems and network solutions face a central problem when the two-way communications networks overload from spikes in use. Broadcast content providers lose the use of their web sites when they are most needed. The two-way networks are often clogged with redundant transfers which are better done through a broadcast mechanism.
Additionally, there are problems in determining when these usage spikes occur. The web site being overloaded is in a poor position to call for help and the prior art does not provide an external mechanism to determine web site hits.
One approach to determining when usage spikes would occur involves the use of packet analysis. However, known applications of packet analysis are invasive of either privacy or traffic flow. The router/bridge/firewall applications often alter not only the packets themselves but control how and where they are, or are not, permitted to be sent. Surveillance of network packet traffic can possibly violate the privacy of those communicating on the network.
Another problem with the state of the art is that two-way communication networks are inherently unreliable because access to any content source and/or broadcast data source can fail from too many access requests. This is a major problem to many content providers. Companies paying for advertising seen by millions of people during peak coverage events, such as sporting events, lose an expensive opportunity. Exactly when the most people are interested in finding out about a product or company, their web-site goes down.
Network nodes, e.g. can also fail for basically the same sort of reasons. Consider what happens if a large number of clients simultaneously try to multiple receive streaming video web sites. The nodes involved are likely to fail for essentially the same reason.
What is needed is a method and/or apparatus for improving a network's ability to withstand concentrated spikes, for example, of web-site hits in the two-way communications part of such networks.